The Political Consequences of Perceived Threat and Felt Insecurity

We draw on data from a national RDD telephone sample of 1549 adult Americans conducted between October 15, 2001 and March 2, 2002 to explore the impact of a need for security on support for national security policies in the aftermath of the 911 terrorist attacks. In past research, an external threat has been assumed to have uniform impact on an affected population, a claim that has met with growing research scrutiny. We advance research on threat through an examination of the political effects of individual differences in one’s ability to feel secure in the aftermath of terrorism, exploring the interaction between perceived threat and felt security. Most Americans reported a sense of security after the 911 attacks. But a sense of insecurity among a minority of Americans coupled with a perceived threat of future terrorism increased support for both domestic and international security policy-the curtailment of domestic civil liberties, tougher visa checks, and support for the war in Afghanistan. Our findings underscore the diverse ways in which individuals react politically to a common external threat. We draw on attachment (Bowlby 1982/1969) and terror management theory (Pyzszcynski et al 2002) to understand the origins of individual differences in felt security. Psychological reactions to terrorism play a pivotal role in understanding public support for government anti-terrorist policies. As Crenshaw (1986, p. 400) argues: “The political effectiveness of terrorism is importantly determined by the psychological effects of violence on audiences.” There are differing psychological reactions to external threat, however, and these reactions shape support of government policies designed to combat terrorism. In our past research we have explored the differing political implications of anxiety, anger, perceived personal threat, and perceived national threat (Huddy et al 1995; Huddy, Feldman and Cassese in print; Huddy et al 2003). In the current study, we take this research one step further to explore not only political reactions to threat but how a sense of security interacts with threat to influence support for national security policies. Terrorists aim to threaten members of a target population directly by perpetrating random acts of violence on people engaged in everyday activities. Research on threats that involve the potential for physical harm such as crime, natural disasters, and violent conflicts provide clear evidence that personal threat increases one’s sense of vulnerability and motivates action designed to minimize personal risk (Browne and Hoyt 2000; Ferraro 1996; Sattler et al 2000; Smith and Uchida 1988). We find similar evidence after 9/11. Individuals living in Queens and Long Island (in relatively close proximity to the World Trade Center) who felt personally threatened by terrorism used more caution in handling their mail, spent more time with their families, delayed or dropped their plans to travel by air, and used public transportation in Manhattan less frequently in the several months after the attacks. In this instance, perceived personal threat motivated cautionary action over and above the effects of any perceived risk of terrorism to Americans more generally (Huddy et al 2002). Threat not only motivates protective behaviors, it also promotes support for protective government policies. Studies conducted to date find a clear relationship between national threat and support for national and domestic security policies (Davis and Silver 2004; Huddy et al 2 2005). In our research on reactions to 9/11, we find that perceived future terrorist threat leads to greater support for an aggressive national security policy, including increased support for the war in Afghanistan (Huddy et al 2005). Threat also heightens support for increased surveillance policies against current and potential Arab immigrants to the United States, and policies that promise increased domestic safety but threaten Americans’ civil liberties. National threat leads to support for punitive action against terrorist groups (Bar-Tal and Labin, 2001, Gordon and Arian 2001; Friedland and Merari 1985). And an adversary’s degree of belligerence and the level of threat posed to US interests increase public support for war and military action against a specific aggressor (Herrmann et al. 1999; Jentleson 1992; Jentleson and Britton 1998; Mueller 1973). Threat clearly increases support for aggressive foreign policy but it is less clear whether this serves a retaliatory function or is seen as a form of protection. We further explore the link between threat and support of national security policy by examining the extent to which it depends on a person’s sense of personal security. Some individuals are able to maintain a sense of security even in the face of threatening events that may undercut their need for protective personal and government policies (Pszczyinski et al 2002; Mikulincer and Shaver 2003; Fraley, Fazzari, Bonanno, & Dekel, 2006.). A Need for Security: Attachment and Terror Management Theories There is broad agreement among social scientists that a sense of security is a basic need in humans and other species (Fraley 2005; Goldberg 2000; Mikulincer and Shaver 2003). Psychologist Abraham Maslow (1943) placed security just above the satisfaction of basic physiological needs in his hierarchy of human needs (and below love and self-actualization). And political scientist Ronald Inglehart (1997), building on Maslow’s work, viewed the fulfillment of basic economic and security needs as a necessary societal precondition to the pursuit of postmaterialist values which emphasize freedom, self-expression and quality of life. 3 Bowlby’s attachment theory (1982/69) is an evolutionary-developmental account of social behavior that posits a need for social proximity to protective others under conditions of threat and danger, implicitly assuming a need to maintain a sense of security under threat. Building on Bowlby’s original insights (1969) on the universality of a human need for attachment as way to deal with insecurity, scholars have theorized about the evolutionary advantage of adult attachment under conditions of threat (Ainsworth, Blehar, Water, and Well, 1978; Mikulincer and Shaver, 2003; Fraley, Brumbaugh, and Marks, 2005; Kirkpatrick, 1998; Tancredy and Fraley, 2006). Several different possibilities have been raised. First, adult attachment could improve human reproductive fitness through the process of kin selection by promoting the protection of those with whom one shares similar genes (Hamilton, 1964). Second, adult attachment could be a simple outgrowth of humans’ protracted developmental period in infancy in which attachment is needed to protect offspring and stimulate complex social and cognitive skills (Bowlby, 1969; Fernald, 1993; Mikulincer and Shaver, 2003; Fraley, 2002). Third, adult attachment may foster long-lasting romantic and parenting relationships which may improve the fitness of one’s offspring (Zeifman and Hazan, 1997; although for alternative views see Belsky, 1999; Chisholm, 1996; Buss and Schmitt, 1996; Schmitt, 2005; cf., Kirkpatrick, 1998). As noted by Sroufe and Waters (1977), the goal of attachment behaviors is to reduce anxiety through an established sense of “felt security”. The attachment system emerges in early infancy, particularly in the context of the caregiver-child relationship, and operates as a functional system organizing interpersonal beliefs throughout development (Bowlby, 1969). While the attachment system is universal, operating in all humans and a host of other organisms (e.g., Fraley et al., 2005), individual differences often emerge from variations in attachment histories. Successful early attachment experiences establish one’s chronic attachment style, operating as what Bowlby (1969) referred to as an “internal working model” of the self and other 4 (Bartholomew and Horowitz, 1991). Repeated episodes of successful attachment behavior give rise to a secure attachment style, which is marked by self-confidence, empathy, and trust in both children and adults (Mikulincer & Shaver, 2003; Brennan & Shaver, 1995; Mikulincer & Shaver, 2001). We are primarily concerned with the difference between secure and insecure attachments in this research, but numerous researchers have investigated the consequences of differing types of insecure attachment styles. Once established, attachment styles are relatively constant across the lifespan, suggesting that they serve as a stable individual difference (Fraley, 2002; Ainsworth, 1991) that may be mediated by distinct neural and hormonal reactions to threat. For instance, Kraemer (1992) found that physiological indicators of stress such as norepinephrine varied depending on whether rhesus monkeys were reared in isolation versus with mothers or peers. And humans with an enduring sense of insecurity release higher levels of glucocorticoids in stressful situations than those with a secure attachment (Goldberg, 2000). A good deal of research on attachment theory has been devoted to the assessment of individual differences in attachment style, and the effects of this style on ongoing romantic relationships. But the theory is broader than that, and has been employed to understand individual differences in coping with stressful events. Typically, researchers examine the impact of attachment style on coping with various threats (for a review, see Mikulincer and Shaver 2003). Some of that research is very relevant to the study of reactions to terrorism. Mikulincer and colleagues (Mikulincer, Florian and Weller 1993) examined the effects of the Gulf War, and Iraqi Scud missile attacks, on Israelis with different attachment styles and found that securely attached individuals perceived lower levels of threat, reported higher levels of self-efficacy, actively sought out social support, and pursued constructive problem solving strategies. Former Israeli prisoners of war with a secure attachment style reported that they recalled positive 5 memories or encounters with others to help cope with their imprisonment, in essence seeking symbolic proximity to internalized attachment figures (Solomon et al 1998). The need for attachment can also be activated by a mortality salience manipulation typically employed in the study of terror management theory (Mikulincer, Birnbaum, Woodis, and Nachamis 2000). Secure individuals tend to react to mortality salience with an increased desire for intimacy (Mikulincer and Florian 2000) and greater willingness to engage in social interaction (Taubman Ben-Ari, Findler, and Mikulincer 2002). In contrast, less secure individuals respond to mortality salience with more severe judgment and punishment of transgressors (Mikulincer and Florian 2000). When taken together, research on attachment theory suggests that a long-standing sense of security derived from a secure attachment style and associated active coping strategies may help to mitigate the negative effects of stressful events. The ability to cope well with stressful events may extend to politics, undermining the need among secure individuals for government policies designed to promote a sense of safety. This possibility has not yet received empirical support but offers an intriguing explanation of how long-standing dispositions interact with events to shape support for government security policies. As a consequence, the theory may help to explain why threat is likely to drive one person to demand more powerful safety and security policies from the federal government while another is able to maintain a sense of personal security in the absence of government action. In a popular competing view of how humans maintain a sense of ongoing security, terror management theory (Pyszczinski et al 2002) provides a related but distinct account of how individuals restore a sense of security in the face of threat, especially existential threats linked to one’s morality. Form this perspective, coping involves adherence to a cultural worldview linked, for example, to religion, moral conduct, or patriotism that boosts self–esteem and often involves the derogation of outsiders who do not share the same view. Terror management researchers 6 have paid less attention to differences in the ways in which individuals restore a sense of security but recent research by Hart and colleagues (2005) suggests that differences in attachment style also moderate responses to existential threats. In their research, insecure individuals were most likely to respond to existential sthreat with more positive ratings of an ingroup member. In this research, we closely examine whether a sense of felt security provides a buffer against threatening events and reduces the need for government action to maintain a sense of safety. We focus on both domestic and international government security policies. In essence, we test whether perceived threat coupled with a feeling of insecurity promotes support for policies that enhance domestic security but may reduce civil liberties, and leads to increased support for overseas military action. Anxiety: A Needed Control In turning to security as an important moderating influence on perceived threat, it is important to distinguish its effects from that of anxiety with which it is related. Our past work demonstrates that personal threat is strongly tied to a sense of heightened anxiety. Individuals who felt anxious after 9/11 experienced higher levels of personal threat and were more likely to live in New York city and have known someone who was killed or injured in the attacks (Huddy et al 2005). This highlights a key facet of anxiety. It is in part situational and affected by ongoing events as made clear by much research on state anxiety (Eysenck 1992). But this stands in marked contrast to a sense of security which, unlike anxiety, is grounded in a long-standing sense of personal safety and is unlikely to be affected by a specific stressful event. Anxiety is a product of a stressful event whereas a sense of security mitigates negative psychological reactions to such an event (Mikulincer and Shaver 2003). Research on attachment theory makes clear that anxiety and a secure attachment style are negatively related. A secure attachment predicts better mental health under stressful circumstances, and decreased symptoms of depression and anxiety (Fraley et al, 2006). A sense 7 of felt security is likely to dampen anxious reactions to the events of 9/11 and so be negatively related to it, but we expect the relationship to be modest because other situational factors such as proximity to the events or the severity of their personal impact also shape a sense of personal threat and concomitant levels of anxiety. We also expect anxiety and security to have differing effects on domestic and national security policies. Recent psychological research demonstrates that threat-induced anxiety tends to elevate risk perceptions and risk aversion (Lerner and Keltner 2000; 2001; Lowenstein et al 2000; Raghunathan and Pham 1999). The link between anxiety and risk aversion has important implications for support of national but not domestic security policies. Anxiety is likely to elevate the perceived risks associated with a given military intervention and decrease support for the deployment of military troops, potentially undercutting a desire for retaliation. But it is unlikely to undercut support for domestic security policies which are not inherently risky. We have confirmed these predictions in our prior research on reactions to 9/11 and the Iraq war (Huddy et al 2005; Huddy, Feldman, and Cassese 2005). In contrast, both domestic and international security policies should provide a sense of security to threatened individuals who lack a strong sense of internal safety. Hypotheses We supplement past findings from research on terrorism with recent psychological insights from attachment and terror-management theories on the importance of felt security in dealing with stressful events. We focus specifically on the degree to which the political effects of threat depend on a sense of insecurity. We contrast the origins of felt security with other reactions to the 911 terrorist attacks, and expect security to be a more long-standing reaction than anxiety and depression that is less tied to perceived threat and the personal consequences of the attacks. Consistent with the predictions of attachment theory, we also expect felt security to moderate the influence of threat on anxiety, and produce a greater sense of trust in fellow 8 Americans. We examine whether a sense of security moderates the desire to bolster self-esteem through an increase in American patriotism and denigration of outsiders, consistent with the expectations of a blended attachment-terror management approach. Finally, we evaluate whether felt security moderates the influence of threat on support of national security policy. Felt security should decrease the influence of perceived terrorist threat on support for domestic security policies. It should also decrease the impact of threat on support for overseas military action. In contrast, insecure individuals who perceive pervasive threat should be most inclined to support both domestic and international security policy.